Project Summary/Abstract There is increasing interest in using immunotherapy or targeted therapy to treat brain metastases, including leptomeningeal disease from cancer. However, it can be challenging to interpret response to treatment or to understand patterns of response or resistance to treatment since serial tissue samples are difficult to obtain from the brain. Thus, there is a critical need to identify noninvasive makers of response or resistance to new therapies being evaluated in this understudied patient population. In this proposal, Aim 1 will combine circulating tumor DNA with sophisticated MRI scans to improve our ability to interpret response of brain metastases patients treated with pembrolizumab, a PD-1 inhibitor. With pembrolizumab, increase in contrast enhancement on MRI can be tumor progression or inflammation from successful re-activation of the immune system with treatment. We hypothesize that by combining a quantitative measure of tumor burden, specifically circulating tumor DNA, with MRI scans, we will be able to disambiguate true progression from pseudoprogression and, moreover, to understand mechanisms of response/resistance. Aim 2 will use circulating tumor DNA to shed light on the clonal evolution of brain metastases in response to targeted therapy, allowing for noninvasive means to understand response and resistance patterns. Aim 3 will address the clinical challenge of accurately determining radiographic response in patients with leptomeningeal disease from cancer. Similar to our hypothesis for parenchymal brain metastases, we expect that adding ctDNA to MRI will improve our ability to more reliably measure LMD response to treatment.